Prompt Catcher

Warns you before sending an email, message, or post that still contains leftover AI prompt or summary text.

Ever paste from AI and forget to delete the part where it says "Want me to make this more concise?" or "This version feels a little tighter"? Prompt Catcher catches exactly that — the leftover AI text that don't belong in the email, message, or post you're about to send. As you compose in Gmail, Outlook on the web, or LinkedIn, if it spots likely AI leftovers, a small bar slides up with the specific flags and a confidence score. Click Details to see what it found, or Dismiss to hide it for that text. You can also click the toolbar icon and paste any text to check it on demand — the manual checker works with text from anywhere. WHAT IT CATCHES • Offers to revise: "Would you like me to shorten this?", "Want me to punch it up?" • Self-commentary: "This version feels tighter", "I've kept it warm but professional" • Input requests: "Paste it and I'll format it", "Just give me the topic and I'll write it" • Assistant boilerplate: "Certainly! Here's a revised version", "As an AI language model" • Leftover placeholders: [Your Name], [Recipient's Name] • Stray Markdown: ## headings, **bold**, code fences PRIVATE BY DESIGN Everything runs on your device. Your text is never sent to any server, never stored, and there is no account or login. The detection is a transparent set of pattern rules, not a cloud service. This is a safety net, not a guarantee — it's heuristic, so it can miss cleverly disguised text or occasionally flag a legitimate phrase. It warns you while you compose rather than blocking send.